A Centuries-Long Delay Between a Paleo-Ice-Shelf Collapse and Grounding-Line Retreat in the Whales Deep Basin, Eastern Ross Sea, Antarctica

University of South Florida Scholar Commons

Marine Science Faculty Publications College of Marine Science

2018

A Centuries-Long Delay between a Paleo-Ice-Shelf Collapse and Grounding-Line Retreat in the Whales Deep Basin, Eastern Ross Sea, Antarctica

Philip J. Bart Louisiana State University

Matthew DeCesare Auburn University

Brad E. Rosenheim University of South Florida, [email protected]

Wojceich Majewski Polish Academy of Sciences

Austin McGlannan The University of Oklahoma

Follow this and additional works at: https://scholarcommons.usf.edu/msc_facpub

Part of the Life Sciences Commons

Scholar Commons Citation Bart, Philip J.; DeCesare, Matthew; Rosenheim, Brad E.; Majewski, Wojceich; and McGlannan, Austin, "A Centuries-Long Delay between a Paleo-Ice-Shelf Collapse and Grounding-Line Retreat in the Whales Deep Basin, Eastern Ross Sea, Antarctica" (2018). Marine Science Faculty Publications. 1292. https://scholarcommons.usf.edu/msc_facpub/1292

This Article is brought to you for free and open access by the College of Marine Science at Scholar Commons. It has been accepted for inclusion in Marine Science Faculty Publications by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. www.nature.com/scientificreports

OPEN A centuries-long delay between a paleo-ice-shelf collapse and grounding-line retreat in the Received: 11 April 2018 Accepted: 17 July 2018 Whales Deep Basin, eastern Ross Published: xx xx xxxx Sea, Antarctica Philip J. Bart1, Matthew DeCesare2, Brad E. Rosenheim3, Wojceich Majewski4 & Austin McGlannan5

Recent thinning and loss of Antarctic ice shelves has been followed by near synchronous acceleration of ice fow that may eventually lead to sustained defation and signifcant contraction in the extent of grounded and foating ice. Here, we present radiocarbon dates from foraminifera that constrain the time elapsed between a previously described paleo-ice-shelf collapse and the subsequent major grounding-line retreat in the Whales Deep Basin (WDB) of eastern Ross Sea. The dates indicate that West Antarctic Ice Sheet (WAIS) grounding-line retreat from the continental shelf edge was underway prior to 14.7 ± 0.4 cal kyr BP. A paleo-ice-shelf collapse occurred at 12.3 ± 0.2 cal kyr BP. The grounding position was maintained on the outer-continental shelf until at least 11.5 ± 0.3 cal kyr BP before experiencing a 200-km retreat. Given the age uncertainties, the major grounding-line retreat lagged ice- shelf collapse by at least two centuries and by as much as fourteen centuries. In the WDB, the centuries- long delay in the retreat of grounded ice was partly due to rapid aggradational stacking of an unusually large volume of grounding-zone-wedge sediment as ice-stream discharge accelerated following ice- shelf collapse. This new deglacial reconstruction shows that ongoing changes to ice shelves may trigger complex dynamics whose consequences are realized only after a signifcant lag.

Te foating ice shelves that fringe Antarctic Ice Sheets (AISs) are an important part of the cryosphere because they slow the ofshore fow of grounded ice. Hence, the presence/absence of an ice shelf greatly infuences the balance between ice-volume accumulation and ablation1. In the present interglacial climate, ice shelves are melting from their upper2,3 and lower surfaces4–7. Te current rapid rate of thinning reduces the buttressing efect of ice shelves8. Te collapse of the Larsen Ice Shelf B on the Antarctic Peninsula in 20029 (Fig. 1a) triggered a near instantaneous accelerated fow of the inland ice streams10. Te associated ice-stream defation is of concern because it may eventually lead to major grounding-line retreat8,11, especially in those areas where marine-based ice overlies a foredeepened seafoor12. Tere is, however, a dearth of paleo-data that constrains the geologic timeframes over which dynamic thinning and defation trigger signifcant and sustained grounding-line retreat afer ice-shelf collapse13,14. Grounding-line retreat is crucial to assessing whether ongoing AIS changes signifcantly contribute to global sea level rise. Tis leads to the following question: how much time elapsed between paleo-ice-shelf collapse and signifcant dynamic response in the form of paleo-grounding-line retreat? Several previous reconstructions have shown that paleo-ice shelves have experienced abrupt collapse15–18. Te interplay between paleo-ice-shelf collapse and grounding-line retreat has not previously been constrained due to general paucity of reliable dates from the sedimentologies and geomorphologic features that record calving-front and grounding-line oscillations. A paleo-perspective of changing ice dynamics in response to ice-shelf collapse is

1Louisiana State University, Department of Geology and Geophysics, Baton Rouge, 70803, USA. 2Auburn University, Department of Geosciences, Auburn, 36849, USA. 3University of South Florida, College of Marine Sciences, St. Petersburg, 33701, USA. 4Institute of Paleobiology, Polish Academy of Sciences, Warszawa, 00-818, Warszawa, Poland. 5The University of Oklahoma, School of Geology and Geophysics, Norman, 73019, USA. Correspondence and requests for materials should be addressed to P.J.B. (email: [email protected])

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 1 www.nature.com/scientificreports/

Figure 1. (A) Map of Antarctica showing ice streams and downstream (foating) ice shelves. Te dashed lines in the continental interiors demarcate drainage areas of the East and West Antarctic Ice Sheets that converge into Ross Sea. Te gray shade shows the paleo-Bindschadler Ice Stream drainage area during the LGM. B = Bindschadler Ice Stream; D = David Glacier; Byd = Byrd Glacier; M = Mercer Ice Stream; W = Whillans Ice Stream; K = Kamb Ice Stream; Ma = MacAyeal Ice Stream; E = Echelmeyer Ice Stream; CF = calving front; RIS = Ross Ice Shelf; GL = grounding line; RFIS = Ronne-Filchner Ice Shelf; LIS = Larsen Ice Shelf with the darker blue area showing the extent of the Larsen Ice Shelf that collapsed in 2002. Te modern-day shelf edge position is shown as a dashed line. (B) Bathymetry of the eastern Ross Sea continental shelf. Te paleo-BIS was confned to the WDB between the Hayes and Houtz Banks. Te light gray rectilinear lines are the locations of seismic data. Te squares are core acquired by Mosola and Anderson24 and the crosses are those described by McGlannan et al.18.

crucial to understanding the potential sensitivity of ice stream discharge to ice shelf thinning19 amidst recent observations that grounding lines are retreating throughout Antarctica20. Here, we present radiocarbon dates of in situ foraminifera from sub-ice-shelf and grounding-line-proximal sediments to constrain the time elapsed between a paleo-ice-shelf breakup and the subsequent major grounding-line retreat in the WDB of eastern Ross Sea. Te development of this retreat chronology specifcally relied on a detailed stratigraphic framework that was presented in two recent studies18,21. Background Reconstructing ice-sheet retreat from the geology of continental shelf trough basins. During the Last Glacial Maximum (LGM), i.e., from 24,000 to 19,000 years ago, mass balance had shifed to positive and the extent of ice fow from East and West Antarctica expanded towards the Antarctic continental shelf edge (Fig. 1a)21–26. Broad and wide bathymetric troughs and banks on the outer continental shelves demonstrate that several fast-fowing erosive ice streams formerly drained the AISs (Fig. 1a,b). In Ross Sea, ice fow converged into fve large ice-streams that reached the outer continental shelf (Fig. 1a). Te subsequent retreat of grounded and foating ice is well recorded in the sedimentology and geomorphology of paleo-ice-stream trough basins that were partly backflled with grounding zone wedges (GZWs) and post-glacial sediment successions21,25–28. Te former locations of the ice-sheet grounding line are best reconstructed from the morphological boundary between a gently, landward-dipping GZW topset (the surface to which the ice stream was grounded) and a steeply seaward-dipping GZW foreset (the basinward-dipping marine depositional surface). In other words, GZWs are essentially subaqueous moraines deposited at the marine termination of the ice stream. Identifying and mapping GZW morphology on seismic profles and bathymetric data is crucial to accurately reconstruct AIS retreat because these features represent the former locations occupied by grounded ice and thus provide une- quivocal evidence that ice streams became stationary, or momentarily re-advanced, during their overall retreat. Early deglacial reconstructions were routinely based on reconnaissance level surveys of the major trough basins24. Te physical reconstructions have continually been improved as more marine data have been acquired. Tese improved stratigraphic frameworks are important because they highlight the locations from which a chronology of ice-sheet and ice-shelf retreat from the outer continental shelves can best be obtained.

WAIS retreat from the WDB outer and middle continental shelf. During the austral summer of 2015, the distribution of sea ice was low and sea state was unusually mild in eastern Ross Sea. Tose favorable conditions allowed acquisition of a large-area (2,500 km2) multibeam swath bathymetry survey, three regional dip-oriented seismic refection profles from the center of the WDB and a transect of strategically-located core during expedition NBP1502B (Figs 1b, 2a). Te WDB was formerly occupied by an expanded Bindschadler Ice Stream, which received ice fow from the central part of the WAIS. Retreat of the Bindschadler Ice Stream from WDB has been recently well-constrained using both detailed analysis of seafoor morphology and core

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 2 www.nature.com/scientificreports/

bathymetric saddle outer continental shelf a KC16 middle continental shelf KC15 KC13/14 KC7 KC8 KC10 KC11 KC12 KC9

water GZW5 ice depth (m) GZW4 ice-shelf collapse RIS GZW7 ice C icebergs from 0 mod l large area ice shelf ice 200 i ZW4

Islan d Roosevelt G e f 3 ic W ce GZW7 ice7 f GZ 2 i ice rapidly retreating GZW GZW1 525 grounding line 6 3 638 4 5 1 675 2 LGM 0510 20 30 40 50km EOL NBP1502b-3 distance from the WAIS divide (km) 1000 1150 1200 1250 1300 1350 1400 400 250 200 150 100 50 0 distance from the WDB continental shelf edge (km) shelf edge

b KC9 KC16 KC15 KC13/14 KC7 KC8 KC10 KC11 KC12 0 0 open-marine 0.4±0.2 3 32(3.2 (17)17) calving front 3 4 4

8 calving fron 8 8.7±0.2 sub-ice-shelf open-marine 9 grounding-line position 9 maintained by rapid sedi- 10 mentation and aggrada- 10

t tional stacking 11 11 200 km rapid grounding line retreat at 11.5 7 11.5±0.3 11.77 ice-shelf breakup 12 12.0 accelerated ice-streaming 6 12 Wheeler Diagram Legend 12.32 5 4 12.6±0.2 12.3±0.3 13 hiatus 13 13.53 14 open-marine sediment 3 14.0±0.3 14 calving fron ice-shelf-breakup bed 14.7 14.7±0.4 15 2 15 sub-ice-shelf sediment sub-ice-shelf 16 7 GZW7 15.5 9 16 1 12.6±0.2 foraminifera age (cal. kyr BP) t 17 17.1 17 13.5 inferred age (cal. kyr BP) 18 18

Figure 2. (A) Line drawing interpretation of regional dip-oriented seismic line NBP1502B_3 with projected core stations shown as vertical lines. Te transect location is shown with a solid bold line in B. At the southern end, we show our projection of stratal relationships below the Ross Ice Shelf. Te gray shaded units labeled 1–7 are GZWs21 with the successive ice sheet and ice shelf positions occupied through time18. Te multiple grounding line and ice shelf positions shown afer GZW7 represent rapid retreat of the grounding line to Roosevelt Island. RISmod = the modern position of the Ross Ice Shelf. (B) Wheeler Diagram corresponding to line drawing interpretation of NBP1502B_3 (shown in A). Te diagram shows distance along the earth’s surface on the horizontal axis. Unlike the seismic section, the vertical axis of a Wheeler Diagram represents geologic time. In other words, any horizontal line through the diagram is an isochron representing lateral changes in paleoenvironmental setting. Te diferent colored zones on the diagram show both the horizontal and vertical changes in sedimentary environments through time as constrained by seismic stratigraphy and seafoor geomorphology as well as sediment facies and radiocarbon ages. Te core stations used to construct the sedimentary facies shown on the diagram are posted on the top of the Wheeler Diagram. Te key foraminiferal radiocarbon dates from this study are posted adjacent to the black vertical lines, whose lengths represent the sediment penetration at the various core stations. Te gray numbers are inferred ages. Te 3.2 kyr date is from Conway et al.47.

sedimentology. Multibeam swath bathymetric surveys coupled with subsurface stratigraphic imaging21 revealed that the paleo-Bindshadler ice stream (paleo-BIS) had advanced to the continental shelf edge as evidence by megascale glacial lineations (MSGLs). A backstepping succession of seven GZWs that partly bury the earlier-formed MSGLs marks the initial retreat of the groundling line from the continental shelf edge. Tis over- lapping stack (GZW1 through 7) indicates that the grounding line paused multiple times as the ice stream vacated the WDB (Fig. 2a)21; four of the seven GZWs are partly exposed on the outer continental shelf of the WDB21. Grounding zone wedges 4 through 7 exhibit an overall aggradational stacking pattern that is manifest as the large bathymetric saddle between the Hayes and Houtz Banks located approximately 70 km from the continental shelf edge (Fig. 1b). In contrast, the lack of GZWs on the middle continental shelf, i.e. south of the bathymetric saddle, suggests that afer the deposition of GZW7, the grounding line abruptly shifed 200 km towards Roosevelt Island (Figs 1b, 2a). Only a series of small morainal ridges that mantle the topset of GZW7 on the middle continental shelf record the major retreat. The favorable conditions during the NBP1502B expedition allowed strategic coring with respect to the exposed GZWs to reconstruct the patterns with which the grounding line and the calving front oscillated between groundings18. Most of the cores penetrated to massive diamict that was deposited either in a subglacial or grounding-line-proximal setting as the grounding line was retreating across the outer continental shelf. The upcore sediment transitions from diamict to glacial marine sediments (representative of sub-ice-shelf and open-marine environments) record the time-transgressive retreat of both grounded and foating ice. Te

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 3 www.nature.com/scientificreports/

Figure 3. SEM images of pristine preservation among calcareous foraminifera from cores recovered during expedition NBP1502B; Trifarina earlandi specimens are shown in the upper row and Globocassidulina biora specimens are shown in the lower row.

correlation of the geophysical and geological data permitted a much more detailed reconstruction of WAIS retreat18 (Fig. 2a) than was possible with previously available reconnaissance-level data from the WDB24. Most importantly, although the grounding line paused seven times during retreat, no major back and forth oscillations of either grounded or foating ice occurred between these groundings18. Te paleo-ice shelf that formerly covered the outer continental shelf collapsed during the deposition of GZW418 (Fig. 2a). Despite ice-shelf collapse, the regional stratigraphic framework requires that the grounding line remained on the outer continental shelf to deposit GZWs 5 through 7 in an overall aggradational stacking pattern18. Grounded ice then abruptly retreated an additional 200 km to Roosevelt Island (on the inner continental shelf) and a large ice shelf reformed over the middle continental shelf (Fig. 2a)18. Te Ross-Ice-Shelf calving front eventually moved south to its modern position. Te combination of detailed surface and subsurface mapping of the WDB with strategic coring allowed by anomalously good sea state and sea ice conditions allows for tremendous insight into the former grounding locations as well as the precise stratigraphic boundaries in sediment cores that record the key transitions in paleo-environmental conditions that record calving-front and grounding-line oscillations.

Foraminiferal record from the WDB. A recent study by Majewski et al.22 found that ice-proximal sediments in the WDB contain relatively abundant in situ foraminifera. Benthic and/or planktonic foraminifera were found in all cores, however, the abundance and species composition difered between and throughout each core. In general, the upper parts of the cores down to 10 to 20 cm below the sediment/water interface represented open-marine conditions and contained agglutinated foraminifera with abundances and diversities that decreased with depth. Agglutinated foraminifera were dominated by Miliammina arenacea associated with Paratrochammina, Reophax, Spiroplectammina and Labrospira22. Below 10 to 20 cm core depth, calcareous foraminifera were present at most locations except from cores on the middle continental shelf (i.e., KC14, KC15, and KC16), which were almost barren of calcareous foraminifera. In contrast, in all underlying sediments on the outer continental shelf, the total foraminiferal abundances were variable but strongly dominated by calcareous foraminifera, reaching in some intervals up to 40 specimens per gram of dry sediment. Taxonomic composition, although limited to less than twenty-fve species in total, was variable22. Te calcareous assemblage was always dominated by Globocassidulina and/or Trifarina, accompanied by Cibicides, Ehrenbergina, Astrononion, Epistominella, and other benthic genera, in addition to the planktonic Neogloboquadrina pachyderma sinistral. Preservation of calcareous foraminifera was variable but mostly ranged from fair to pristine. Some broken or etched tests were encountered but well-preserved, complete, transparent tests with no sediment infll were domi- nant in the most fossiliferous intervals. Some examples of pristine preservation with intact fragile spines, no signs of wall dissolution, and no secondary flling of very fne pores were present (Fig. 3).

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 4 www.nature.com/scientificreports/

Methods Te detailed understanding of regional geomorphology21, sedimentology18 and foraminiferal presence22 outlined above is an essential framework to our study of retreat chronology in the WDB. With respect to our objective of constraining the time elapsed between paleo-ice-shelf collapse and major grounding-line retreat in the WDB, the stratigraphic framework shows that there are two key sedimentary facies transitions that have to be dated. Te frst transition is the ice-shelf collapse during GZW4 that is recorded in core on the outer continental shelf by the sedimentary change from the sub-ice-shelf to ice-shelf-breakup units. Te second transition is the subsequent major grounding line retreat at the end of GZW7 that is recorded in core from the GZW7 foreset by the sedimentary change from diamict to open-marine diatom ooze. Our methodology focused on radiocarbon dating of well-preserved foraminifera isolated from near these two sedimentary boundaries. In instances where foraminifera were not sufciently abundant for 14C analysis near the facies boundary, the closest interval below the sedimentary facies boundary was used.

Foraminiferal selection. Sediment samples (150 cm3 volume) were collected from all kasten cores (KC) at intervals of 5 to 10 cm for foraminiferal analyses directly afer onboard visual core description. Care was taken to not include abrupt facies transitions in a single sample. Te jumbo piston cores (JPCs) were cut and sampled using a pair of 50 cm3 sediment plugs in July 2015 afer their arrival at the Antarctic Marine Geology Research Facility in Tallahassee, Florida. Sediment samples were collected every 10-cm using 2 cm thick core intervals. Samples directly below and above each facies transition were processed using the methods outlined below. Tese samples did not always yield enough biogenic carbonate in pristine condition for radiocarbon analysis. In these instances, preliminary foraminiferal abundances from Majewski et al.22 were used as a reference for sample selection. Sediment was washed through a 63 μm sieve to remove silt and clay fractions and dried at ~50 °C. Sediment was then dry-sieved using 149, 250 and 425 μm fractions before collecting foraminifera. Tests were picked under a binocular light microscope and placed into microslides for archiving. Specimens were identifed to the species level. Only unbroken tests that did not exhibit any signs of discoloration breakage were isolated using a binocular light microscope and used in further analysis. A minimum of 0.6 mg of calcite was measured for radiocarbon analysis.

Radiocarbon analyses. Foraminiferal samples were sent to the University of South Florida College of 14 Marine Science to be converted to CO2, along with modern and C-free carbonate reference materials, by reaction with phosphoric acid (H3PO4) in vacuo. Glass ampoules of CO2 gas were sent to Woods Hole Oceanographic Institute to be analyzed for 14C on the National Ocean Sciences Accelerator Mass Spectrometer. One sample of G. biora was split into two equal masses and replicate ages were within 50 14C yr of one another. Tese data were calibrated to calendar year BP using MARINE13 calibration curve in CALIB 7.129,30 with a marine reservoir efect of 1,300 ± 100 yr (difering by 900 yr from the global average reservoir age of 400 yr) established for Antarctic marine carbonates31. All foraminiferal radiocarbon ages are in cal kyr BP as the median age ± the age range representing the 95% confdence interval as per radiocarbon reporting conventions32. All data generated or analyzed during this study are included in this published article. Results Radiocarbon ages of foraminifera in the WDB. Foraminifera dated for this study came from intervals that showed the most abundant calcareous foraminifera22. Where possible, we used monospecifc samples comprised of benthic G. biora and T. earlandi (and in one instance planktonic N. pachyderma) that were the most numerous and fairly massive (Table 1). Te good to pristine preservation of these foraminifera, along with the lack of clearly reworked specimens, that could be identifed either by diferent coloration, sediment flling33 and/ or presence of extinct species34, indicates that practically all specimens could be considered as being in situ22,35. Rare specimens showing signs of etching, found especially among Cibicides or foraminiferal specimens in cores from the middle continental shelf, although most likely also being in situ, were not used for radiocarbon analysis. We obtained a total of eighteen foraminiferal radiocarbon ages (Table 1). Seventeen of the new radiocarbon dates presented here were generated from benthic foraminifera. Te calibrated age uncertainties are low with ffeen of the eighteen dates having uncertainties ranging from 197 to 346 years (analytical uncertainties ranged between 30–150 14C y). Te largest uncertainty, from the smallest sample mass, was 512 years. Eleven dates were obtained from sub-ice-shelf sediments that accumulated on the outer continental shelf prior to the ice-shelf collapse. Only two of the median calendar ages from the sub-ice-shelf unit on the outer continental shelf are out of stratigraphic order. Given the possible range of calibrated ages represented by the median dates, however, the ages are considered to be in good stratigraphic order. Te oldest date from the base of sub-ice-shelf sediments (in KC9) is 14.7 ± 0.4 cal kyr BP. Te youngest date from the top of the sub-ice-shelf sediments (in KC8) is 12.3 ± 0.3 cal kyr BP. To better constrain the range of timing for sub-ice-shelf sedimentation on the outer continental shelf, two samples from KC12, comprised of mixed benthic species G. biora and T. earlandi, were radiocarbon dated. Tese ages are 13.5 ± 0.2 cal kyr BP at 38 cm depth and 13.8 ± 0.3 cal kyr BP at 97 cm depth. Tese data essentially overlap with the monospecifc age 14.0 ± 0.3 at 53 cm and therefore also represent viable radiocarbon ages for sub-ice-shelf sedimentation on the outer continental shelf prior to ice-shelf collapse. No radiocarbon dates were obtained for the ice-shelf-breakup unit. Only one radiocarbon date, 0.4 ± 0.2 cal kyr BP, was obtained from the upper part of the open-marine sediment on the outer continental shelf at KC03. Four radiocarbon dates were obtained from foraminifera (exclusively G. biora) that were isolated from diamict sediments that accumulated on GZW7 foreset on the outer continental shelf. Te ages come from three core sites that are separated by 50 km. In the one core with multiple dates, i.e., KC05, the calendar ages are in stratigraphic order. Te age uncertainties range from 281 to 416 years. Te calendar ages range from 12.4 ± 0.3 cal kyr BP to 11.5 ± 0.3 cal kyr BP. No radiocarbon dates were obtained for open-marine sediments that overly the

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 5 www.nature.com/scientificreports/

Interval Age Err Calibratd ±age range Sample # Core ID (cm) Lithology Species 14 C yr BP ± Fm error δ13C yr BP range range − + *OS-126095 KC12 38–42 SIS mixed benthic 12,950 65 0.001700 −0.11 13,504 221 13,283 13,742 OS-122390 KC12 53–58 SIS T. earlandi 13,405 120 0.002833 −0.58 13,990 256 13,734 14,246 *OS-126094 KC12 97–102 SIS mixed benthic 13,200 75 0.001800 0.09 13,763 252 13,511 14,015 OS-122389 KC12 215–220 diamict G. biora 44,837 4,200 0.002265 −0.29 46,651 512 46,139 47,163 OS-122357 KC11 80–85 A SIS G. biora 12,597 70 0.001808 −0.1 13,177 223 12,954 13,400 OS-122358 KC11 80–85B SIS G. biora 12,644 70 0.001809 −0.11 13,221 217 13,004 13,438 OS-131664 KC11 80–85 SIS N. pachyderma 12,550 35 0.00090 0 13,130 228 12,880 13,335 OS-122363 KC11 180–185 SIS G. biora 13,201 80 0.001920 −0.56 13,765 252 13,513 14,017 OS-122364 KC10 100–105 SIS G. biora 13,162 80 0.001915 0.36 13,720 245 13,475 13,965 OS-122394 KC09 30–35 SIS T. earlandi 11,702 150 0.004432 −0.56 12,268 319 11,949 12,587 OS-122368 JPC09 235–237 SIS G. biora 13,797 85 0.001925 0.31 14,683 449 14,234 15,132 OS-122391 KC08 40–45 SIS T. earlandi 12,010 110 0.003136 −0.51 12,630 202 12,428 12,832 OS-122388 JPC07 235–237 diamict G. biora 11,664 85 0.002532 0.54 12,214 346 11,868 12,560 OS-122360 KC06 146–148 diamict G. biora 11,818 75 0.002115 0.04 12,421 281 12,140 12,702 OS-122385 KC05 64–69 diamict G. biora 11,268 90 0.002725 0.65 11,472 341 11,131 11,813 OS-122359 JPC05 280–282 diamict G. biora 11,432 70 0.002116 0.73 11,740 416 11,325 12,156 OS-122383 KC03 20–25 OM N. pachyderma 1,618 35 0.003302 0.68 357 197 160 554 OS-131665 KC16 200–204 SIS mixed benthic 8,660 30 0.00130 0.47 8,223 201 8,001 8,403

Table 1. Radiocarbon ages for foraminifera. Sample numbers are from NOSAMS. Te asterisks symbol denotes samples that are mixed benthic species. 14 C yr BP is uncalibrated ages. Calibrated yr BP is the median age results from Calib 7.1 sofware with ± age range being the 95% confdence interval. SIS = sub-ice shelf; OM = open marine.

GZW7 diamict because those sediments contained only a few calcareous foraminifera. Te new radiocarbon dates (Table 1) are shown with respect to the lateral sedimentary facies changes on a Wheeler Diagram (Fig. 2b). Discussion A total of eleven radiocarbon dates from six core sites (KC12, −11, −10, −9, −8 and −6; see Table 1) constrain the timing of sub-ice-shelf sedimentation on the WDB outer continental shelf prior to the paleo-ice-shelf collapse. Te oldest radiocarbon date from sub-ice-shelf sediment, 14.7 ± 0.4 cal kyr BP, comes from core station KC09 (Fig. 2a,b). Its stratigraphic location atop diamict deposited on the GZW2 foreset indicates that a third grounding-line shif away from the continental shelf edge (i.e., the establishment of a grounding line associated with GZW3 with its adjacent sub-ice-shelf environment) had occurred by 14.7 ± 0.4 cal kyr BP, (Fig. 2a,b; Table 1). Likewise, the geologic principles of stratigraphic superposition and facies relationships (in particular, the existence of sub-ice-shelf sediments above the GZW3 foreset at core station JPC/KC08, Fig. 2a,b) require that a small ice shelf continuously covered the outer continental shelf by the time that the grounding line had experienced a fourth shif away from the continental shelf edge (as recorded by shif from GZW3 to GZW4)18. Te foraminiferal radiocarbon dates indicate that the ice shelf remained in existence until at least 12.3 ± 0.2 cal kyr BP (Fig. 2b) but its collapse occurred shortly thereafer and is recorded by an ice-shelf-breakup event bed (Fig. 2a,b)18. Te youngest date from sub-ice-shelf sediments, 12.3 ± 0.2 cal kyr BP, is from immediately below the ice-shelf-breakup unit in KC8. Te area of the ice shelf that collapsed was likely similar to that of the Larsen B Ice Shelf that collapsed in 200236,37. Based on the regional stratigraphic framework, the loss of ice-shelf buttressing was coeval with the end of GZW418 (Fig. 2b). Afer ice-shelf collapse, a grounding-line ice clif would have existed at the grounding line (Fig. 2b). To sustain such a confguration, a rapid increase in ice stream fow38,39 and an increase in sediment flux would have accompanied it. In the WDB, this is recorded by a significant change from backstepping (GZW1 through −4) to progradational/aggradational (GZW5 through −7, see Fig. 2a). Our radiocarbon ages (Table 1) provide strong support that the signifcantly larger volume GZWs (GZW5 through −7, see Fig. 2a,b) were the result of rapid deposition that would have been associated with increased ice stream fow. In the WDB, the grounding-line position on the outer continental shelf was apparently maintained by rapid subglacial and grounding-zone sedimentation40 as represented by the aggradational stacking of GZW4 through −7 which flled the accommodation created by ice-stream defation from accelerated ice-stream fow21,41. A total of fve radiocarbon dates from four core sites (KC3, −5, −6 and −7; see Table 1) constrains the timing of diamict deposition on the GZW7 foreset. Our estimated date for the major grounding line retreat that followed paleo-ice-shelf collapse is based on the youngest age of benthic foraminifera from GZW7 diamict, 11.5 ± 0.3 cal kyr BP. Tis date is taken from immediately below the sedimentologic boundary that marks the abrupt transition to an open-marine setting in KC7. Bart et al.35 showed that calcareous foraminifera from GZWs can be used for radiocarbon dating of marine diamicts from the eastern Ross Sea if they are selected with considerable care. Taken at face value, the median radiocarbon ages for paleo-ice-shelf collapse and major grounding-line retreat (i.e., 12.3 and 11.5 cal kyr BP, respectively) suggest that the grounding-line position on the WDB outer continental shelf was maintained for eight centuries afer the ice shelf collapse. Fully considering the calibration-induced age

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 6 www.nature.com/scientificreports/

ranges, the grounding-line position persisted for a minimum of 200 years and a maximum of 1400 years afer ice shelf breakup. In a geologic context, even our maximum estimate could be considered an instantaneous response. As concerns human-relevant time scales, an important point here is that the ongoing changes to ice shelves may trigger complex dynamics whose consequences are realized only afer a signifcant lag. Te land-based glaciers that were formerly fronted by the Larsen Ice Shelf B experienced immediate acceleration following the ice-shelf collapse in 2002 however, there has of yet been no signifcant retreat in the grounding line. In the case of the WDB, a centuries-long lag necessitates sediment aggradation from a high sediment fux. Sediment fux to the grounding line depends on several factors including the ice-sheet drainage area42 as well as the substrate erodibility and the presence/absence of melt water43,44. In the case of the WDB, easily erodible unconsolidated Plio-Pleistocene strata underlie the WDB45 and hence the paleo-BIS sediment fux was sufciently high to deposit an unusually large-volume compound GZW41 that helped maintain a relatively stationary grounding line. In the WDB, the subsequent 200-km retreat of the grounded line toward Roosevelt Island constituted an abrupt additional 15% contraction in the extent of grounded ice. A large-area ice shelf reformed over the middle continental shelf (Fig. 2a,b; see the upcore transitions from subglacial till to sub-ice-shelf sediments at KC13 thru KC16). A radiocarbon date from the middle continental shelf (at KC16) indicates that this second ice shelf was still in existence by 8.6 ± 0.2 cal kyr BP (Fig. 2a,b). Compound specifc radiocarbon dates suggest that the calving front had migrated south of the middle continental shelf before 5 kyr BP46 but there is no sedimentologic evidence that a second catastrophic breakup of the reformed ice shelf that covered the middle continental shelf. Grounded ice eventually retreated past Roosevelt Island at 3.2 kyr BP47. Tus far, all consideration of uncertainty of the timing of paleo-BIS deglaciation has incorporated analytical uncertainty (relatively small) and calibration uncertainty (larger). Certainly, the sub-ice-shelf melting, if compa- rable to today, could possibly have been driven at least in part by a water mass with an older age (greater reservoir age). Deep-sea stylasterid hydrocorals record radiocarbon age diferences during Circumpolar Deep Water (CDW) incursion of about 100–200 years (King et al., in review). Tis is less than, but a signifcant portion of, the minimum amount of lag between ice shelf breakup and grounding line retreat that our foraminifera dates constrain. Given that the reservoir age of warmer water masses (e.g., today’s CDW) is related to far feld processes that create it in the north Atlantic as well as local processes that pull it onto the continental shelf, it is difcult to postulate the potential for a large diference in reservoir age during deglaciation compared to what is observed today. Indeed, there is evidence of a slowdown of North Atlantic Deep Water formation, which would potentially have increased reservoir ages during the LGM48. However, the overall properties of Southern Ocean water masses also involve more local processes related to the Antarctic Circumpolar Current and shelf-slope exchange. Tese are less well constrained, limiting how much we can forecast diferences in the reservoir age of subsurface, potentially warm water masses that could have driven the ice dynamics preserved in WDB sediments. Te only previous study of the WAIS retreat from the eastern Ross Sea outer continental shelf relied on bulk acid-insoluble organic matter isolated from deglacial sediment24. Teir core stations in the WDB are not ideally located and hence they are not useful for our study. More importantly, they considered their radiocarbon dates suspect because the bulk organic matter yielded very old ages at the sediment-water interface. Apparently, older organic carbon is mixed with that generated in the water column contemporaneous with deglacial sedimentation, a prognosis that is typical in Antarctic margin sediments49. Our results confrm that a highly-detailed retreat chronology from in situ foraminifera can be obtained via synthesis of regional-scale geophysical surveys and targeted sediment coring. A combination of factors (collapse of the ice shelf that covered the outer continental shelf, accelerated fow and ice-stream defation, as well as the aggradation of a foredeepened profle by the end of GZW7 deposition) all predisposed the paleo-BIS to its subsequent rapid major (200 km) retreat to Roosevelt Island at 11.5 ± 0.3 cal kyr BP (Fig. 2b). Te negative mass balances associated with ice-shelf collapse, accelerated ice-stream discharge and eventual grounding-line shif would have delivered freshwater to the Southern Ocean with the potential to alter thermohaline circulation and other aspects of the global climate system50. A review of the literature on Antarctic margins indicates that many ice streams experienced backstepping retreats of foating and grounded ice, but, given uncertainties in calibrations for diferent dating strategies, it is not yet possible to defnitely say whether the WDB ice-shelf collapse and grounding-line shifs were synchronous with ice-sheet changes recorded elsewhere25,51–55. Te WDB ice-shelf collapse may have been synchronous with an increase in iceberg rafed debris fux in Scotia Sea referred to as AID2 by Weber et al.53. Teir composite records of ice-rafed debris from the Scotia Sea permits the possibility that either pulsed ice-shelf collapse, subsequent accelerated ice-stream discharge or abrupt grounding-line retreat may have ultimately caused the increased iceberg rafing fuxes53. Calving-front and grounding-line shifs are signifcant measures of climate change. Te geological perspective provided by these results are important to climate models that seek to predict dynamic grounding-line and calving response to either external and/or internal forcing50,56,57. Te new detailed reconstruction from the WDB shows that a paleo ice-shelf collapse triggered complex dynamics whose consequences were not fully realized until afer a lag of at least two centuries. Even in isolation, the ice volumes returned to the global ocean from the WDB in the time elapsed since the paleo-ice-shelf collapse at 12.3 ± 0.2 cal kyr BP and subsequent 200-km grounding-line retreat at 11.5 ± 0.3 cal kyr BP were relevant to coastal settings in terms of the resulting sea-level rise and coastal transgressions it would have produced. As concerns the potential for future ice-shelf collapse in the present-day interglacial, it is worth noting that many ice-stream systems are now grounded on the foredeepened inner continental shelves that is presumably underlain by either basement rock or other less erodible substrate. Hence, a transition to unbuttressed fow of those modern ice streams might be less able to maintain a grounding position by accelerated deposition of sediment within the grounding zone.

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 7 www.nature.com/scientificreports/

References 1. Alley, R. B. et al. Oceanic Forcing of Ice-Sheet Retreat: West Antarctica and More. Annual Review of Earth and Planetary Sciences 43, 307–231 (2015). 2. Trusel, L. D., Frey, K. E., Das, S. B., Munneke, P. K. & van den Broeke, M. R. Satellite-based estimates of Antarctic surface meltwater fuxes. Geophysical Research Letters 40, 6148–6153, https://doi.org/10.1002/2013GL058138 (2013). 3. Kingslake, J., Ely, J. C., Das, I. & Bell, R. E. Widespread movement of meltwater on and across Antarctic ice shelves. Nature 544, 349–352 (2016). 4. Rignot, E. & Jacobs, S. S. Rapid bottom melting widespread near Antarctic Ice Sheet grounding lines. Science 296, 2020–2023 (2002). 5. Marcott, S. A. et al. Ice-shelf collapse from subsurface warming as a trigger for Heinrich events. Proceedings of the National Academy of Science 103(33), 13415–13419 (2011). 6. Paolo, F. S., Fricker, H. A. & Padman, L. Volume loss from Antarctic ice shelves is accelerating. Science 348, 327–331 (2015). 7. Noel, G. et al. Channelized melting drives thinning under a rapidly melting Antarctic Ice Shelf. Geophysical Research Letters 44, 9796–9804, https://doi.org/10.1002/2017/GL074929 (2017). 8. Rignot, E., Mouginot, J., Morlighem, M., Seroussi, H. & Scheuchl, B. Widespread, rapid grounding line retreat of Pine Island, Twaites, Smith, and Kohler glaciers West Antarctica, from 1992 to 2011. Geophysical Research Letters 41, 3502–3509 (2014). 9. Banwell, A. F., MacAyeal, D. R. & Sergienko, O. V. Breakup of the Larsen B Ice Shelf triggered by chain reaction drainage of supraglacial lakes. Geophysical Research Letters 40, 5872–5876 (2013). 10. Glasser, N. F. & Scambos, T. A. A structural glaciological analysis of the 2002 Larsen B ice shelf collapse. Journal of Glaciology 54(184), 3–16 (2008). 11. Hughes, T. J. Te “weak underbelly” of the West Antarctic Ice Sheet. Journal of Glaciology 27, 518–525 (1981). 12. Joughin, I. & Alley, R. B. Stability of the West Antarctic Ice Sheet in a warming world. Nature Geoscience 4, 507–513 (2011). 13. Joughin, I., Smith, B. E. & Medley, B. Marine ice sheet collapse potentially under way for the Twaites Glacier Basin, West Antarctica. Science 344, 735–738 (2014). 14. Scambos, T. A., Bell, R. E., Alley, R. B., Anandakrishnan, S. & Bromwich, D. H. Brunt et al. How much, how fast? A science review and outlook for research on the instability of Antarctica’s Twaites Glacier in the 21st century. Global and Planetary Change 153, 16–34 (2017). 15. Kilfeather, A. A. et al. Ice-stream retreat and ice-shelf history in Marguerite Trough, Antarctic Peninsula: Sedimentological and foraminiferal signatures. Geological Society of America Bulletin 123(5-6), 997–1015 (2011). 16. Kirshner, A. E. et al. Post-LGM deglaciation in Pine island Bay, West Antarctica. Quaternary Science Reviews 38, 11–26 (2012). 17. Witus, A. E. et al. Meltwater intensive glacial retreat in polar environments and investigation of associated sediments: example from Pine Island Bay, West Antarctica. Quaternary Science Reviews 85, 99–118 (2014). 18. McGlannan, A. J., Bart, P. J., Chow, J. M. & DeCesare, M. On the influence of post-LGM ice shelf loss and grounding zone sedimentation on West Antarctic Ice Sheet Stability. Marine Geology 392, 151–169 (2017). 19. Reese, R., Gudmundsson, G. H., Levermann, A. & Winkelmann, R. Te far reach of ice-shelf thinning in Antarctica. Nature Climate Change 8, 53–57 (2018). 20. Konrad, H. et al. Net retreat of Antarctic glacier grounding lines. Nature Geoscience 11, 258–262 (2018). 21. Bart, P. J., Anderson, J. B. & Nitsche, F. O. Post-LGM grounding-line positions of the Bindschadler Ice Stream in Ross Sea. Journal of Geophysical Research: Earth Surface 122, 1827–1844 (2017). 22. Majewski, W., Bart, P. J. & McGlannan, A. J. Foraminiferal assemblages from ice-proximal paleo-settings in the Whales Deep Basin, eastern Ross Sea, Antarctica. Palaeogeography, Palaeoclimatology, Palaeoceangraphy 493, 64–81 (2018). 23. Domack, E. W., Jacobson, E. A. & Anderson, J. B. Late Pleistocene – Holocene retreat of the West Antarctic Ice-Sheet system in the Ross Sea: Part 2 - Sedimentologic and stratigraphic signature. Geological Society of America Bulletin 111(10), 1517–1536 (1999). 24. Mosola, A. B. & Anderson, J. B. Expansion and rapid retreat of the West Antarctic Ice Sheet in eastern Ross Sea: possible consequence of over-extended ice streams? Quaternary Science Reviews 25(17–18), 2177–2196, https://doi.org/10.1016/j.quascirev.2005.12.013 (2006). 25. Smith, J. A. et al. Deglacial history of the West Antarctic Ice Sheet in the western Amundsen Sea Embayment. Quaternary Science Reviews 30(5–6), 488–505, https://doi.org/10.1016/j.quascirev.2010.11.020 (2011). 26. Anderson, J. B. et al. Ross Sea paleo-ice sheet drainage and deglacial history during and since the LGM. Quaternary Science Reviews 100, 31–54, https://doi.org/10.1016/j.quascirev.2013.08.020 (2014). 27. Bart, P. J. & Cone, A. N. Early stall of West Antarctic Ice Sheet advance on the eastern Ross Sea middle shelf followed by retreat at 27,500 14CyrBP. Palaeogeography, Palaeoclimatology, Palaeoecology 335–336, 52–60, https://doi.org/10.1016/j.palaeo.2011.08.007 (2012). 28. Klages, J. P., Lindow, J. & ÓCofaigh, C. Reconstruction of changes in the Amundsen Sea and Bellingshausen Sea sector of the West Antarctic Ice Sheet since the Last Glacial Maximum. Quaternary Science Reviews 100, 55–86 (2014). 29. Stuiver, M., Reimer, P., & Reimer, R. CALIB 7.1: http://calib.qub.ac.uk/calib/ (2005). 30. Reimer, P. J. et al. IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years cal BP. Radiocarbon 55(4), 1869–1887, https://doi.org/10.2458/azu_js_rc.55.16947 (2013). 31. Berkman, P. A. et al. Circum-Antarctic coastal environmental shifs during the Late Quaternary refected by emerged marine deposits. Antarctic Science 10(3), 345–362, https://doi.org/10.1017/S0954102098000406 (1998). 32. Millard, A. Conventions for Reporting Radiocarbon Determinations. Radiocarbon 56(2), 555–559, https://doi.org/10.2458/56.17455 (2014). 33. Majewski, W. & Anderson, J. B. Holocene foraminiferal assemblages from Firth of Tay, Antarctic Peninsula: Paleoclimate implications. Marine Micropaleontology 73(3–4), 135–147, https://doi.org/10.1016/j.marmicro.2009.08.003 (2009). 34. Webb, P.-N. P. & Strong, C. C. P. Foraminiferal biostratigraphy and palaeoecology in Upper Oligocene–Lower Miocene glacial marine sequences 9, 10, and 11, CRP-2/2A drill hole, Victoria Land Basin, Antarctica. Palaeogeography, Palaeoclimatology, Palaeoecology 231(1–2), 71–100, https://doi.org/10.1016/j.palaeo.2005.07.036 (2006). 35. Bart, P. J., Coquereau, L., Warny, S. & Majewski, W. In situ foraminifera in grounding zone diamict: a working hypothesis. Antarctic Science 28(4), 313–321 (2016). 36. Rack, W. & Rott, H. Pattern of retreat and disintegration of the Larsen B ice shelf, Antarctic Peninsula. Annals of Glaciology 39(1), 505–510 (2004). 37. Scambos, T. et al. Detailed ice loss pattern in the northern Antarctic Peninsula: widespread decline driven by ice front retreats. Te Cryosphere 8(6), 2135–2145 (2014). 38. Scambos, T. A., Bohlander, J. A., Shuman, C. A. & Skvarca, P. Glacier acceleration and thinning afer ice shelf collapse in the Larsen B embayment, Antarctica. Geophysical Research Letters 31(18), 2001–2004, https://doi.org/10.1029/2004GL020670 (2004). 39. Rignot, E. et al. Accelerated ice discharge from the Antarctic Peninsula following the collapse of Larsen B ice shelf. Geophysical Research Letters 31(18), 2–5, https://doi.org/10.1029/2004GL020697 (2004). 40. Alley, R. B., Anandakrishnan, S., Dupont, T. K., Parizek, B. R. & Pollard, D. Efect of sedimentation on ice-sheet grounding-line stability. Science 315, 1838–1841 (2007). 41. Bart, P. J., Krogmeier, B. J., Bart, M. P. & Tulaczyk, S. M. Te paradox of a two millennia grounding duration of the West Antarctic Ice Sheet in Ross Sea. Scientifc Reports 7, 1–8, https://doi.org/10.1038/s41598-017-01329-8 (2017b).

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 8 www.nature.com/scientificreports/

42. Bart, P. J. & Owolana, B. On the duration of West Antarctic Ice Sheet grounding events in Ross Sea during the Quaternary. Quaternary Science Reviews 47, 101–115 (2012). 43. Hallet, B., Hunter, L. & Bogen, J. Rate of erosion and sediment evacuation by glaciers: a review of feld data and their implications. Global and Planetary Change 12, 213–235 (1996). 44. Elverhoi, A., Hooke, R. L. & Solheim, A. Late Cenozoic erosion and sediment yield from the Svalbard-Barents Sea region: implications for understanding erosion of glacierized basins. Quaternary Science Reviews 17, 209–241 (1998). 45. Hayes, D. & Frakes, L. General synthesis deep sea drilling project leg 28. Initial Reports of the Deep Sea Drilling Project 28, 919–942 (1975). 46. Yokoyama, Y. et al Widespread collapse of the Ross Ice Shelf during the late Holocene. Proceedings of the National Academy of Sciences 1–6, https://doi.org/10.1073/pnas.1516908113 (2016). 47. Conway, H., Hall, B. L., Denton, G. H., Gades, A. M. & Waddington, E. D. Past and Future Grounding-Line Retreat of the West Antarctic Ice Sheet. Science 280, 280–284, https://doi.org/10.1126/science.286.5438.280 (1999). 48. Rosenheim, B. E. et al, Antarctic sediment chronology by programmed-temperature pyrolysis: Methodology and data treatment. Geochemistry, Geophysics, Geosystems 9, https://doi.org/10.1029/2007GC001816 (2008). 49. McManus, J. F., Francois, R., Gheradi, J.-M. & Keigwin, L. D. Collapse and rapid resumption of Atlantic meridional circulation linked to deglacial climatec changes. Nature 428, 834–837 (2004). 50. Golledge, N. R. et al. Antarctic contribution to meltwater pulse 1A from reduced Southern Ocean overturning. Nature Communications 5, https://doi.org/10.1038/ncomms6107 (2014). 51. Licht, K. J. Te Ross Sea’s contribution to eustatic sea level during meltwater pulse 1A. Sedimentary Geology 165(3–4), 343–353, https://doi.org/10.1016/j.sedgeo.2003.11.020 (2004). 52. Heroy, D. C. & Anderson, J. B. Radiocarbon constraints on Antarctic Peninsula Ice Sheet retreat following the Last Glacial Maximum (LGM). Quaternary Science Reviews 26(25–28), 3286–3297, https://doi.org/10.1016/j.quascirev.2007.07.012 (2007). 53. Weber, M. E. et al. Millennial-scale variability in Antarctic ice-sheet discharge during the last deglaciation. Nature 510(7503), 134–138, https://doi.org/10.1038/nature13397 (2014). 54. Hall, B. L., Denton, G. H., Heath, S. L., Jackson, M. S. & Kofman, T. N. B. Accumulation and marine forcing of ice dynamics in the western Ross Sea during the last deglaciation. Nature Geoscience 8(8), 625–628, https://doi.org/10.1038/ngeo2478 (2015). 55. Fogwill, C. J. et al. Antarctic ice sheet discharge driven by atmosphere-ocean feedbacks at the Last Glacial Termination. Scientifc Reports 7, https://doi.org/10.1038/srep39979 (2017). 56. Church, J. A. et al. Sea Level Change. In: Climate Change. Te Physical Science Basis. Contribution of Working Group I to the FifhAssessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, 1137–1216 (2013). 57. DeConto, R. & Pollard, D. Contribution of Antarctica to past and future sea-level rise. Nature 531(7596), 591–597, https://doi. org/10.1038/nature17145 (2016). Acknowledgements Tis project was supported by NSF grant #1246357 to P.J.B. We thank the science and shipboard crew from expeditions NBP1502B and -A. Author Contributions P.J.B. conceived the experiment. M.D. and W.M. isolated and identified the in situ foraminifera that were radiocarbon dated. M.D. and B.E.R. generated the radiocarbon ages and calibrations. P.J.B., M.D. and W.M. acquired the geologic and geophysical data during expedition NBP1502B. A.M. and P.J.B. conducted the sedimentologic analyses. Te fgures were drafed by P.J.B. and M.D. All authors contributed to the interpretation of the data. All authors reviewed the manuscript. Additional Information Competing Interests: Te authors declare no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. Te images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Scientific REPOrTS | (2018)8:12392 | DOI:10.1038/s41598-018-29911-8 9